Re: [lojban] NORATS, SPACE, and PUBLIC in PEG grammar

[".alyn.post." <alyn.post@lodockikumazvati.org>]

On Tue, Nov 23, 2010 at 11:02:15AM -0800, Robin Lee Powell wrote:
> On Tue, Nov 23, 2010 at 11:57:35AM -0700, .alyn.post. wrote:
> > I settled on extending the grammar definition by tagging
> > expressions, which pollutes the grammar but makes the code using
> > the grammar easier to write.  It doesn't accomplish your goal of
> > not having a bunch of non-grammar garbage in the peg file.  :-(
> 
> Well, show me what you end up with when you're done and we'll see
> how it looks.
> 

It doesn't look stellar:

http://bugs.call-cc.org/browser/release/4/genturfahi/trunk/tests/mex.peg

I choose to use Racket Scheme-style keyword tags, which have the
form #:FOO.  Note also that this grammar uses Scheme-style
character expressions, like #\+ and #\*, which are ugly in a way
that doesn't affect the Lojban grammar, which would express the same
as [+] and [*].


> what Lisp PEG parser generator are you using?  :)
> 
> -Robin
> 

Jerk.  :-)

I'm writing one:

  http://wiki.call-cc.org/eggref/4/genturfahi

Last night I finished the boostrap parser generator, and I'm
comparing differences in my hand-written parser generator vs
the one it is generating.  I of course have issues to fix
before my parser generator is self-hosting, but I'm down around 10
differences in the bootstrap generator vs the one written in PEG,
and the last several differences have been bugs in the bootstrap
rather than bugs in the version the bootstrap is generating.

I started by using the only scheme PEG parser I knew about, but I
quickly gave up on it and realized I needed something that would
work directly with PEG files:

  http://wiki.call-cc.org/eggref/4/genturfahi#history

I managed to find two bugs in Chicken Scheme while I was writing the
parser, both related to DSSSL-style #!rest, #!optional, and #!key
parameters, so I suspect I'm the only person on the planet who has
tried using them:

  https://bugs.call-cc.org/ticket/421
  https://bugs.call-cc.org/ticket/430

As well as learn something about Scheme I didn't know:

  https://bugs.call-cc.org/ticket/433

The bootstrap compiler is compiling the morphology and morphology
header file, but I'm still working on the peg grammar itself.

I've written the code with the idea that I could port it to other
scheme systems, particularly Racket, though this is only theoretical
at this point, as there is zero demand for such a thing.  :-)

Given that Lojban is used as an example of a complex PEG grammar:

  http://en.wikipedia.org/wiki/Parsing_expression_grammar#External_links

I'm not sure it's a bad idea to have a peg parser generator written
specifically to parse Lojban.  I do wish there had been something
available already, but I'm not aware of Scheme code that parsers PEG
files--they all seem to want to write the grammar definition in
Scheme itself.

-Alan
-- 
.i ko djuno fi le do sevzi

-- 
You received this message because you are subscribed to the Google Groups "lojban" group.
To post to this group, send email to lojban@googlegroups.com.
To unsubscribe from this group, send email to lojban+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/lojban?hl=en.